Somya Kansal

I’m an AI-focused, tech-savvy student driven by turning data into decisions. I’m eager to contribute as a Data Scientist, building models and products that improve accuracy, stability, and real user outcomes.

In my Data Analytics Intern role at TechtitudeTribeServicePrivateLimited, I cleaned, preprocessed, and performed exploratory data analysis on large datasets using Python (Pandas, NumPy), leading to a 20% increase in data accuracy and a 25% improvement in business insights.

As a MachineLearning Intern at VerveBridge, I developed and evaluated supervised machine learning models for placement prediction, achieving 25% model accuracy and increasing system stability by 20%. I enjoy designing end-to-end pipelines—from data prep to model evaluation.

My projects reflect my hands-on approach: I built an AI-powered medical chatbot using Python, RAG, Langchain, NLP, Pinecone, and HuggingFace LLM with FastAPI integration. I also created an AI chat application with Streamlit, speech-to-text, and ChatOllama, plus a NutriSmart diet website using Machine Learning (Pytorch, Tensorflow, CNN) and ComputerVision—driving a 25% increase in user engagement.